.png)
ABBYY Vantage Integration Guide
ABBYY Vantage integrates with enterprise systems through OAuth-secured REST APIs that submit documents, track asynchronous processing, and return classification and extraction results.
ABBYY Vantage integration options at a glance
ABBYY Vantage primarily integrates through REST APIs for OAuth-authenticated document submission, skill selection, processing-status tracking, and retrieval of classification and extraction results. File and attachment handling is central because documents are uploaded for asynchronous processing and returned artifacts may be retrieved through the API. Some deployments may expose callback or notification options for selected processing events, but universal webhook coverage was not confirmed, so polling is the safer default. Martini can receive files or API requests, submit them to Vantage, persist processing identifiers, orchestrate polling, normalize results, apply confidence and validation rules, and deliver approved data to downstream applications.
| Integration point | Supported by ABBYY Vantage? | Common use cases | How Martini supports it |
|---|---|---|---|
| REST APIs | Yes | Authenticate API clients, submit documents, select skills, check processing status, and retrieve classification and extraction results. | Martini can consume the Vantage REST API, generate reusable API assets from its definition where available, and orchestrate calls in workflows. |
| File / attachment APIs | Yes | Upload document files for classification and extraction and retrieve supported processed files or artifacts. | Martini can receive files from APIs, repositories, email, SFTP, or other sources, then pass content and metadata to Vantage. |
| Authentication | Yes | OAuth-based client authentication with deployment-specific token endpoints, permissions, scopes, and tenant settings. | Martini can store client credentials and environment-specific URLs as secrets and use bearer tokens when calling Vantage. |
| Bulk / async / batch APIs | Limited | Vantage processing is asynchronous and supports tracking individual processing requests; a separate general-purpose bulk endpoint was not confirmed. | Martini can implement controlled batching, queues, correlation IDs, bounded concurrency, retries, and result aggregation. |
| Webhooks / outbound callbacks | Limited | Some deployments may expose callbacks or notifications for selected processing events, but universal event coverage was not confirmed. | Martini can receive a documented callback through an API workflow; otherwise it can use scheduled polling as the default pattern. |
| GraphQL APIs | Not confirmed | No official ABBYY Vantage GraphQL API was confirmed. | Martini should use the documented Vantage REST interface rather than assuming GraphQL support. |
| SOAP APIs | Not confirmed | No current Vantage SOAP interface was confirmed. | Martini can consume SOAP generally, but SOAP should not be selected for Vantage without deployment-specific confirmation. |
| Database / analytics access | No | Direct access to the internal Vantage database is not an appropriate general public integration mechanism. | Martini should use Vantage APIs and exported results instead of querying internal Vantage tables. |
How ABBYY Vantage exposes data and business events
ABBYY Vantage REST APIs
ABBYY Vantage exposes REST APIs for client authentication, document submission, skill invocation, processing-status checks, and retrieval of classification and extraction results. Exact endpoint paths and payloads should be confirmed against the Swagger or OpenAPI definition for the target deployment.
Martini implementation pattern
Martini implementation pattern: A workflow obtains or reuses an OAuth access token, submits the document and selected skill, persists the returned processing identifier, retrieves the completed result, and maps the output to a canonical or downstream model.
Implementation sequence
ABBYY Vantage file processing
Document files are central to Vantage integration. APIs accept files for classification and extraction, while supported processed files or artifacts may be retrieved according to the deployment's documented behavior, file limits, and retention settings.
Martini implementation pattern
Martini implementation pattern: An upstream API, repository, email flow, SFTP process, or file source provides the document to a Martini workflow, which preserves filename and MIME metadata, submits the content to Vantage, and routes the normalized result.
Implementation sequence
ABBYY Vantage asynchronous processing
Vantage processing should generally be treated as asynchronous because classification and extraction may outlast a normal request-response exchange. A separate general-purpose bulk API was not confirmed, although multiple processing requests can be coordinated.
Martini implementation pattern
Martini implementation pattern: Martini uses a scheduled workflow, queue, or documented callback to advance each request, applies bounded concurrency and retry rules, and aggregates completed results while keeping failed or low-confidence items in an exception path.
Implementation sequence
Common ABBYY Vantage integration patterns
Pattern 1: Extract invoices and purchase orders
When to use this pattern
Use this pattern when finance documents arrive from a repository, email intake, SFTP process, or business application and must be converted into validated supplier, invoice, tax, total, and line-item data. The workflow should distinguish low-confidence extraction from technical failure.
Integration direction
Example Mapping
| ABBYY Vantage Field | Canonical Field | Target Field |
|---|---|---|
| supplier name | supplierName | Supplier |
| invoice number | invoiceNumber | Invoice Number |
| invoice total | totalAmount | Gross Amount |
| line items | lineItems | Invoice Lines |
Martini implementation pattern
Martini receives the file, submits it with the appropriate Vantage document skill, polls the processing request, and normalizes the extraction result. Validation checks required fields, confidence thresholds, date and amount formats, tax calculations, and duplicate business keys before writing to SAP S/4HANA or NetSuite. Transient API failures are retried, while invalid documents and low-confidence results are routed for review.
Martini capabilities used
- workflows
- API consumption
- file handling
- data mapping
- business rules
- error handling
- scheduled execution
Pattern 2: Route claims and supporting documents
When to use this pattern
Use this pattern when a claims application receives forms, correspondence, receipts, and supporting documents that require classification before being routed to different claims processes. It is useful when each document type has different extraction and validation rules.
Integration direction
Example Mapping
| ABBYY Vantage Field | Canonical Field | Target Field |
|---|---|---|
| classification result | documentType | Claim document category |
| claim identifier | claimId | Claim ID |
| extraction result | extractedFields | Claim attributes |
| confidence information | confidence | Review status |
Martini implementation pattern
A Martini API accepts the document and correlation ID, submits it to Vantage, and tracks the asynchronous transaction. Classification determines the downstream route, while confidence and required-field rules determine whether the document proceeds or enters an exception queue. The workflow records the Vantage request ID to support reconciliation and duplicate prevention.
Martini capabilities used
- APIs
- workflows
- orchestration
- data mapping
- conditional routing
- queues
- error handling
Pattern 3: Process onboarding and identity documents
When to use this pattern
Use this pattern when customer or employee onboarding applications need names, addresses, dates, document identifiers, or registration details extracted from uploaded documents. It supports controlled validation before data is written to a system of record.
Integration direction
Example Mapping
| ABBYY Vantage Field | Canonical Field | Target Field |
|---|---|---|
| name | personName | Name |
| address | postalAddress | Address |
| document identifier | documentNumber | Document Number |
| classification result | documentType | Document Type |
Martini implementation pattern
Martini receives the upload, chooses the configured Vantage skill, and polls for completion. It transforms dates and addresses, validates required identity fields and confidence values, and checks the source document ID or content hash before updating Salesforce or ServiceNow. Failed validation is separated from authentication, transport, and Vantage processing errors.
Martini capabilities used
- API exposure
- API consumption
- file handling
- mapping and transformation
- validation
- business rules
- duplicate prevention
Pattern 4: Expose a centralized document-processing API
When to use this pattern
Use this pattern when multiple internal applications need a consistent document-processing contract without each application implementing Vantage authentication, polling, and result normalization independently.
Integration direction
Example Mapping
| ABBYY Vantage Field | Canonical Field | Target Field |
|---|---|---|
| source-system identifier | sourceId | Consumer source ID |
| document type | requestedDocumentType | Normalized document type |
| processing or transaction ID | processingId | Status reference |
| extraction result | normalizedResult | Consumer payload |
Martini implementation pattern
Martini exposes an API that accepts document content, source ID, requested type, and correlation ID, then immediately returns an accepted response. A workflow submits the document to Vantage and manages polling or a documented callback. A status endpoint exposes normalized results, while reusable validation, retry, audit, and exception logic remains centralized.
Martini capabilities used
- API exposure
- workflows
- asynchronous orchestration
- data transformation
- reusable services
- authentication
- monitoring
Applications commonly integrated with ABBYY Vantage
ABBYY Vantage can be integrated with document repositories, business applications, and automation platforms that need classification or extracted field data. These are standards-based integration patterns; a dedicated ABBYY Vantage connector is not required.
| Application | Scenario | Direction | Martini Pattern |
|---|---|---|---|
| Salesforce | Extract information from customer forms, applications, correspondence, or supporting documents and update Salesforce records. | Salesforce → Martini → ABBYY Vantage → Salesforce | Martini receives a Salesforce request or document, submits it to Vantage with the selected skill, polls for completion, validates extracted fields, and updates the relevant Salesforce object. |
| ServiceNow | Process service requests, forms, invoices, and attachments, then populate incidents, requests, or custom records with extracted data. | ServiceNow → Martini → ABBYY Vantage → ServiceNow | A Martini API or workflow accepts the ServiceNow attachment, correlates the Vantage processing request, applies confidence and validation rules, and writes approved values back to ServiceNow. |
| SAP S/4HANA | Extract invoice, purchase-order, delivery, or supplier data before validation or posting in SAP business processes. | SAP S/4HANA → Martini → ABBYY Vantage → SAP S/4HANA | Martini obtains documents from an SAP-connected intake process or repository, submits them to Vantage, maps extraction results to SAP fields, and routes failed validation to an exception workflow. |
| Oracle NetSuite | Automate accounts-payable document capture and map extracted invoice data into vendor bills and related objects. | Oracle NetSuite → Martini → ABBYY Vantage → Oracle NetSuite | Martini orchestrates document submission and asynchronous status checks, transforms invoice and line-item data, validates totals, and writes eligible results to NetSuite. |
| Microsoft SharePoint | Use SharePoint as a document intake or repository location and enrich documents with extracted metadata. | Microsoft SharePoint → Martini → ABBYY Vantage → Microsoft SharePoint | A Martini workflow receives or retrieves a SharePoint file, submits it to Vantage, and updates document metadata or routes the enriched artifact after processing. |
| Microsoft Dynamics 365 | Process customer, supplier, or finance documents and update Dynamics or Dataverse data with extracted fields. | Microsoft Dynamics 365 → Martini → ABBYY Vantage → Microsoft Dynamics 365 | Martini maps Vantage classification and extraction results to Dynamics entities, applies required-field and confidence rules, and handles duplicate or failed submissions. |
| UiPath | Combine Vantage document understanding with robotic process automation for downstream applications that lack suitable APIs. | UiPath → Martini → ABBYY Vantage → UiPath | Martini provides the API orchestration layer for document intake, Vantage processing, normalized results, and status callbacks or polling used by UiPath automations. |
| Microsoft Power Automate | Trigger document-processing workflows from Microsoft 365 events and deliver extracted values to business processes. | Microsoft Power Automate → Martini → ABBYY Vantage → Microsoft Power Automate | Power Automate invokes a Martini API, Martini manages Vantage submission and asynchronous completion, and the normalized result is returned to the Microsoft 365 process. |
How to build a ABBYY Vantage integration in Martini
Objective
Configure the exact ABBYY Vantage deployment URL, OAuth token endpoint, client registration, permissions, and selected skill settings for each environment.
Instructions in Martini
- Store client credentials, URLs, tenant values, and skill identifiers as environment-specific secrets or configuration.
- Confirm the Vantage API version, OAuth requirements, scopes, and file limits for the deployment.
- Use least-privilege permissions and avoid placing credentials or document contents in workflow definitions or logs.
Objective
Select the event, API request, file arrival, or schedule that starts document processing.
Instructions in Martini
- Use a Martini API when an upstream application submits a document.
- Use a scheduled workflow for polling when Vantage callbacks are unavailable or not confirmed.
- Use a queue or controlled schedule for higher-volume processing.
Objective
Receive or retrieve the document, preserve source metadata, and submit it to Vantage with the appropriate document skill.
Instructions in Martini
- Validate file type, size, MIME type, and required source identifiers.
- Call the Vantage REST API with the OAuth bearer token.
- Persist the Vantage processing or transaction identifier before continuing.
Objective
Manage asynchronous processing without assuming that the initial submission contains final extraction data.
Instructions in Martini
- Poll the status endpoint with bounded retries and exponential backoff.
- Use a documented callback only when the target Vantage deployment supports the required notification.
- Apply concurrency limits and retain correlation IDs for reconciliation.
Objective
Transform classification and extraction results into a stable internal or target-system model.
Instructions in Martini
- Map fields, tables, line items, dates, numbers, and confidence information.
- Apply required-field, confidence, total, tax, and business-key validation rules.
- Route low-confidence or inconsistent results to human review or an exception workflow.
Objective
Deliver approved extraction results and supported artifacts to downstream applications while preventing duplicate writes.
Instructions in Martini
- Map the normalized result to the target application's object model.
- Use source IDs, hashes, or business keys for idempotency.
- Record the Vantage request ID, skill, status, timestamps, and final disposition.
Common ABBYY Vantage data objects used in integrations
| Object | Typical Use | Common target systems | Martini handling |
|---|---|---|---|
| Document | Source file submitted for classification and extraction, with metadata, processing state, and associated results. | SharePoint, Salesforce, ServiceNow, SAP S/4HANA, Oracle NetSuite | Martini preserves source identifiers and metadata, submits the file through the REST API, and tracks its processing lifecycle. |
| Document skill | Configured Vantage capability that determines classification or extraction behavior for a document type. | Vantage administration, document intake workflows, downstream business applications | Martini selects the configured skill from environment-specific configuration and records the selected skill for auditability. |
| Classification result | Document type or class assigned by Vantage to route subsequent extraction or business processing. | Claims platforms, Salesforce, ServiceNow, SAP S/4HANA | Martini maps the class to routing rules and selects downstream transformations or exception paths. |
| Extraction result | Structured values extracted from a document, including confidence, validation status, locations, tables, or line items. | SAP S/4HANA, Oracle NetSuite, Salesforce, Microsoft Dynamics 365 | Martini normalizes fields, validates confidence and business rules, transforms types, and writes approved results. |
| Processing request or transaction | Asynchronous operation created when Vantage accepts a document for processing. | Martini state store, queues, monitoring systems, source applications | Martini persists the identifier, correlation data, status, retry count, and completion timestamps for polling and reconciliation. |
| Project or tenant configuration | Tenant-level users, roles, skills, and processing settings used primarily for administration and deployment automation. | Vantage environments, deployment configuration stores | Martini treats deployment URLs, permissions, skills, and tenant settings as environment-specific configuration and secrets. |
Authentication and security considerations
OAuth and client credentials
ABBYY Vantage API clients generally authenticate with OAuth-based access tokens issued by the target deployment. Client registrations, permissions, scopes, tenant settings, token endpoints, and base URLs can vary between cloud, private-cloud, and on-premises environments.
Secrets and least privilege
- Store client credentials, token settings, deployment URLs, and tenant values in Martini Secrets Management.
- Use the minimum Vantage permissions required for document submission, status checks, and result retrieval.
- Do not log access tokens, document contents, or sensitive extracted fields.
Operational considerations for ABBYY Vantage integrations
Asynchronous state
Persist the Vantage processing or transaction ID, source document ID, correlation ID, selected skill, status, retry count, and timestamps. Do not assume document submission returns the final result.
Polling, limits, and retries
- Use bounded polling with exponential backoff when callbacks are unavailable.
- Confirm limits for concurrent submissions, uploads, polling, downloads, file size, and retention.
- Retry transient network and service failures, but avoid repeatedly retrying invalid credentials, invalid skills, or unsupported files.
Data quality and change control
- Validate required fields, confidence values, dates, amounts, totals, taxes, and line items before downstream writes.
- Track skill and API-version changes and normalize results into a stable internal contract.
- Use source IDs, hashes, or business keys to prevent duplicate submissions and writes.
Why use Martini instead of scripts or point-to-point integrations?
Centralized orchestration
Martini centralizes authentication, document submission, asynchronous status handling, result retrieval, transformation, validation, routing, and exception handling instead of duplicating this logic across scripts or point-to-point integrations.
Reusable integration assets
Teams can expose a controlled API for internal consumers, reuse workflows and mappings, and adapt Vantage results to multiple target applications without coupling every system directly to Vantage.
Operational reliability
- Apply consistent retries, backoff, idempotency, correlation, and audit tracking.
- Keep deployment-specific credentials and endpoints outside workflow logic.
- Use monitoring, logs, testing, and controlled deployment practices for document-processing integrations.
Frequently asked questions
ABBYY Vantage can be integrated primarily through OAuth-secured REST APIs. An integration submits document files, selects a document skill, tracks the asynchronous processing request, and retrieves classification and extraction results for delivery to downstream applications. Some deployments may provide callbacks for selected events, but polling should be treated as the default unless callbacks are documented for the target environment.
Yes. Martini can consume the ABBYY Vantage REST API, submit files, track processing requests, retrieve results, expose APIs for upstream applications, and map validated extraction data into downstream systems. A dedicated native Martini connector was not verified.
No. A dedicated ABBYY Vantage connector is not required. Martini can use Vantage's confirmed native REST APIs, OAuth authentication, file submission capabilities, and deployment-specific callback options, with scheduled polling as the conservative asynchronous pattern.
Lonti does not charge an additional per-connector or per-vendor fee to integrate ABBYY Vantage. The integration is subject to the provisioned capacity of the Martini environment. Separate ABBYY, infrastructure, storage, processing, or other third-party costs may apply depending on the deployment and usage model.
REST APIs and file submission are the primary methods for current integrations. Use OAuth-based authentication, submit documents with the appropriate skill, track asynchronous processing, and retrieve results through the documented API for the exact Vantage version and deployment. No official GraphQL or current SOAP API was confirmed.
Some Vantage versions or configurations may expose callbacks or notifications for selected processing events, but a universal all-event webhook model was not confirmed. Martini can receive a documented callback through an API workflow; scheduled polling is the safer default when callback support is unavailable or uncertain.
Synchronization normally starts when a document is received from an application, repository, email process, SFTP flow, or API. Martini submits it to Vantage, stores the processing or transaction identifier, polls or receives a documented completion notification, retrieves the result, validates it, and writes the normalized data to target systems. Source IDs and hashes can prevent duplicate submissions.
Martini can separate authentication, invalid-request, unsupported-file, processing, validation, network, and result-retrieval failures. Transient failures can use bounded retries and backoff, while invalid documents and low-confidence results can be routed to exceptions. Persisting the source ID, content hash, correlation ID, and Vantage request ID supports idempotency and reconciliation.
Related Martini documentation
Transformation
Integrate ABBYY Vantage with Martini
Use Martini to connect ABBYY Vantage document processing with enterprise applications through secure APIs, asynchronous workflows, reusable mappings, and controlled operational processes.